In recent years, mobile phones have become the de facto system of communication across the planet. Mobile phones have helped increase economic growth and critical response in many parts of the world. Mobile phones are even being used for data transmission. However, little academic research has been done on the specific problem of streaming real time video originating from the cameras of mobile devices over cell phone networks. There are many factors that complicate this problem including the limited computational resources of mobile phones, the low and variable bandwidth of cell phone networks, and the need for video compression and streaming algorithms that can be supported by both the mobile phones and cell phone networks. This chapter examines the problems involved and discusses on-going research on the topic. The main goal of this chapter is to identify the real time constraints and challenges of compressing and streaming video from mobile devices for the purpose of designing efficient video compression and streaming techniques that are able to work within the constraints of the limited computational resources and bandwidth available to mobile devices.
TopIntroduction
Due to the proliferation of mobile communication devices, mobile phones and cell phone networks are rapidly becoming the de facto system of communication and even data transmission in many parts of the world. Just a few years ago in many developing regions of the world, access to a telephone was considered a luxury, but today mobile phones are a common aspect of everyday life. Privatized cell phone networks have succeeded in extending international communications access to regions where state phone monopolies had failed to provide even basic service. As an extreme example, in Somalia cell phone companies have extended communications services to the populace over the past decade, despite the collapse of the national government in 1992 (Leeson, 2007).
Contemporary mobile phones have the potential to do much more than old style telephones since they are in reality light computing devices. Aside from the obvious use as telephones, they offer a number of other computational and telecommunication capabilities. In support of communication, for instance, people store address and phone books, calendars and notes on their mobile phones. Thousands of other applications are available, including the ability to access the World Wide Web. Additionally, many mobile phones are capable of producing and viewing high resolution photos and videos. It is only a matter of time before mobile phone users come to expect the ability to stream real time video from mobile phones.
The advantages of streaming real time video from mobile phones are explored in the next section of this chapter. The timeliness of this topic is illustrated by the newest Apple iPhone (Sutter, 2010), which can capture video and transmit it in real time to other iPhones. Most mobile phones today only capture video, store the video in a data file, and then require that the file be transmitted over the cell phone network to a recipient to be played once the entire transmission has been completed, as opposed to a live video stream. The iPhone is able to accomplish the live streaming video only if it is transmitted over Wireless Fidelity (WiFi).
There are significant challenges that must be dealt with to realize the capability of transmitting live streaming video over cell phone networks. First is the inherently low and variable bandwidth of cell phone networks. Cell phone network performance is impacted by a number of independent problems with the potential for dropping numerous packets. Because the size of a raw video stream greatly exceeds the average bandwidth capacity of cell phone networks, the video must undergo some form of compression prior to transmission. The second challenge is that mobile phones and other mobile devices lack computing resources to compress raw video frames efficiently in real time. These devices have processors that compare with desktop/laptop processors of 5-10 years ago, limited memories, and battery management systems which further reduce computing power as the battery drains. The necessity of compressing raw video in real time for streaming requires video compression algorithms that can execute quickly and yet the limited and variable bandwidth of the cell phone networks require that each video frame be reduced in data size to something that the cell phone network will not drop. And so the ultimate challenge is to identify video compression and streaming techniques that can accomplish the task amid these difficult challenges while still delivering a video of watchable quality.
There are several video compression and streaming algorithms. These algorithms, designed for use on desktop/laptop computers, have computational demands that are not suited to the processors of mobile phones. Therefore, these algorithms cannot simply be ported to mobile phones. These ideas will be explored in the chapter.